All-AI-Tools - AI Directories Tool

Overview

All-AI-Tools is an open-source Retrieval-Augmented Generation (RAG) system that indexes and serves information about a curated catalog of over 300 AI tools. According to the GitHub repository, the project combines LlamaIndex for document indexing and retrieval, pgvector-backed PostgreSQL for vector similarity search, and FastAPI to expose query endpoints. A Telegram bot front end lets users query the indexed tool database conversationally, while Nvidia NeMo Guardrails are used to constrain and sanitize model outputs for safer, policy-driven responses. The architecture is focused on lightweight, reproducible RAG workflows for discovery and Q&A about AI products and services. The repository targets developers and researchers who want a self-hosted directory and conversational interface: it ships components for indexing tool metadata, performing vector searches, and serving RAG answers via HTTP and Telegram. According to the project page, it is released under the MIT license and aims to be a practical reference implementation of combining LlamaIndex, pgvector, and safety tooling like NeMo Guardrails.

GitHub Statistics

• Stars: 5
• Contributors: 1
• License: MIT
• Primary Language: Python
• Last Updated: 2024-11-09T15:57:06Z

According to the GitHub repository, All-AI-Tools has 5 stars, 0 forks, and 1 contributor. The project is MIT licensed and shows recent activity with the latest commit on 2024-11-09T15:57:06Z. The small contributor count and low star/fork numbers indicate a small, early-stage project with limited community engagement; however, the recent commit shows it is actively maintained by its author. There are no large community signals (many PRs, forks, or external integrations) visible from the repository metadata.

Installation

Install via pip:

git clone https://github.com/jmrobles/all-ai-tools.git

cd all-ai-tools

python -m venv venv

source venv/bin/activate  # or .\venv\Scripts\activate on Windows

pip install -r requirements.txt

Ensure PostgreSQL has the pgvector extension enabled (CREATE EXTENSION IF NOT EXISTS vector;)

export DATABASE_URL="postgresql://user:pass@localhost:5432/all_ai_tools"

uvicorn main:app --host 0.0.0.0 --port 8000 --reload

Key Features

• Indexes metadata for over 300 AI tools for discovery and Q&A
• RAG pipeline built on LlamaIndex for retrieval-augmented answers
• Vector similarity search using pgvector in PostgreSQL
• FastAPI endpoints to serve RAG responses over HTTP
• Telegram bot interface with Nvidia NeMo Guardrails for safer replies

Community

Community engagement is currently limited: the repository has 5 stars, 0 forks, and a single listed contributor. The project appears actively maintained by its author (latest commit 2024-11-09), but there is minimal external contribution, issues, or discussion visible from the repository metadata. Contributors who want to help should open issues or PRs on the GitHub repo and follow the MIT license terms.

Key Information

• Category: Directories
• Type: AI Directories Tool